Instrumentalism About Practical Reason: Not by Default ∗ Thomas Schmidt

Philosophical Explorations, 2016 Vol. 19, No. 1, 17–27, http://dx.doi.org/10.1080/13869795.2015.1134632

Instrumentalism about practical reason: not by default ∗ Thomas Schmidt

Department of Philosophy, Humboldt-Universita¨t zu Berlin, Berlin, Germany (Received 25 April 2014; ﬁnal version received 12 December 2015)

Instrumentalism is the view that all requirements of practical reason can be derived from the instrumental principle, that is, from the claim that one ought to take the suitable means to one’s ends. Rationalists, by contrast, hold that there are requirements of practical reason that concern the normative acceptability of ends. To the extent that rationalists put forward these requirements in addition to the instrumental principle, rationalism might seem to go beyond instrumentalism in its normative commitments. This is why it is sometimes thought that rationalism is stronger than instrumentalism in a way that entails that instrumentalism is the default view, while rationalists carry the burden of proof. In this paper, I explore and discuss different ways of spelling out this idea. I argue that rationalism is not stronger than instrumentalism in a way that has implications for matters of justiﬁcation and differences in prima facie defensibility of the two sorts of views. Keywords: instrumentalism; instrumental principle; rationalism; practical reason; normativity

1. Instrumentalism vs. rationalism Instrumentalism is the view that all requirements of practical reason can be derived from the instrumental principle, that is, from the following claim:

(IP) one ought to take the suitable means to one’s ends.1,2

When it comes to spelling out a defensible version of instrumentalism, complications

Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 abound. Cases of morally outrageous or otherwise seemingly unacceptable ends and cases of ends the realisation of which would frustrate the agent are among the stumbling blocks for instrumentalists that any reasonably detailed formulation of instrumentalism needs to be able to deal with. In response to such problems, instrumentalists have come up with a number of reﬁnements of their view that are subject to discussion in the literature. Even though it is controversial on grounds of considerations about normative acceptability whether a convincing conception of instrumentalism can be worked out, there seem to be, at least at ﬁrst sight, important asymmetries in philosophical ambition between instrumentalism and rationalism, its main theoretical competitor. Rationalism is thought to be more ambitious to the extent that it involves a commitment to some version of the instrumental principle, as does instrumentalism, but goes beyond what is common ground between the two types of views. Whereas instrumentalists hold that all requirements of practical reason can be derived from (IP), rationalists of the sort just

∗Email: [email protected]

# 2016 Taylor & Francis 18 Thomas Schmidt

characterised believe that in addition to (IP), there are substantial requirements of practical reason that concern the normative acceptability of ends.3 Both instrumentalism and rationalism should be understood as involving the view that actions that are neither required nor prohibited by reason are rationally permitted. Versions of rationalism that also uphold the instrumental principle (in addition to the substantial criteria normatively constraining the set of ends it is rationally permissible to pursue) would seem to be committed to farther reaching and, thus, stronger, claims than instrumentalism in a way which entails a theoretical advantage for instrumentalists. Con- sider, for example, the following remarks by Nozick:

Instrumental rationality is within the intersection of all theories of rationality (and perhaps nothing else is). In this sense, instrumental rationality is the default theory, the theory that all discussants of rationality can take for granted, whatever else they think. [ ...]. The instrumental theory of rationality does not seem to stand in need of justiﬁcation, whereas every other theory does. (Nozick 1993, 133)4

If Nozick and those who share this view were right, then there would be a theoretically highly signiﬁcant asymmetry between instrumentalism and rationalism due to purely structural differences between the two types of view. However, the question whether such an asymmetry obtains has not been discussed in the literature in its own right. As I show in this paper, advances in the philosophical discussion about practical reason and normativity have interesting consequences for it. I argue that there is no sense of “stronger” in which rationalism is stronger than instrumentalism in a way that entails that rationalists are in greater need of justifying their view than instrumentalists in virtue of structural differences between the two views. In arguing that rationalism is not stronger than instrumentalism due to structural differences between the two sorts of theories, I focus on a sense in which rationalism might seem to be theoretically more ambitious than instrumentalism, and which is furthermore different from other such senses that are often discussed in the literature. Many believe that rationalism is more ambitious, or indeed false, due to the type of ought claims that rationalists, as opposed to instrumentalists, bring into play. Some hold, for instance, that rationalism is more ambitious than instrumentalism on the purported grounds that the instrumental principle can better be accommodated within a naturalist framework than characteristically rationalist ought claims. The traditional point of reference for the view that rationalists

Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 have a harder time explaining how their principles are tenable in the ﬁrst place is, of course, Kant’s claim that categorical as opposed to hypothetical imperatives require speciﬁc philosophical underpinning. Others go further and believe that judgements about categorical requirements of practical reasons that are not contingent on antecedent desires or ends of an agent are all false. This is sometimes argued for with reference to the alleged “queerness” of categorical normative facts or of the epistemic faculty which one would need to assume in order to explain how knowledge about such facts is possible (Mackie 1977, 38–42). Others bring into play considerations about the structure and motivational force of practical deliberation (Williams 1979). Against such arguments, Korsgaard (1986) and Hampton (1998, ch. 4) have aimed at showing that there is no difference in type between the normative judgements that rationalists, as opposed to instrumentalists, are committed to maintain that would make instrumentalism more acceptable from a naturalistic point of view. It is controversial whether Korsgaard and Hampton have succeeded in establishing this claim, in particular in view of the recent forceful critique of their arguments by Schroeder (2007, 46–50, 56–83). Instrumentalism About Practical Reason: Not by Default 19

However, since the discussion provided in this paper takes up the question whether instrumentalism is to be seen as the default theory due to purely structural differences between instrumentalism and rationalism, its focus is a different sort of asymmetry claim than the one at issue between Korsgaard and Hampton on the one hand and their opponents on the other. A further asymmetry claim that is not directly related to the discussion to follow is the view, defended by Dreier and others, that the instrumental principle has “a kind of sine qua non status” (Dreier 2001, 43), that is, that the instrumental principle needs to be assumed as a principle of practical reason if anything does, and that it is the only principle with this status. Even if this were so, however, this would only amount to a defence of the instrumental principle, not of instrumentalism – and it is the dialectical situation concerning the latter that is at issue in this paper. (Consider, as an analogy, the claim that no soup not containing salt is seasoned appropriately – containing salt, then, is a sine qua non condition for acceptable soups. Even if this were so, it would not follow that soups only seasoned with salt should be seen as default, whereas all other ways of seasoning soups, involving salt but also other things – say, pepper – face a greater justiﬁcatory burden.) In what follows, I treat the dispute between instrumentalists and rationalists as a contro- versy about what one ought (or has reason) to do. I conceive this issue, in turn, as the subject matter of the theory of practical reason, or practical rationality (following many others, I use these terms interchangeably). Setting up things in this way has been the standard practice in practical philosophy at least until the late 1990s.5 In recent literature, however, many invoke a narrower notion of rationality according to which rationality is only concerned with the internal coherence of mental states, and regard it as an open question how this notion of rationality is related to oughts and reasons.6 Those who understand the notion of practical rationality in this narrow sense should see instrumentalism and rationalism as I discuss them here as theories about practical normativity – and not as theories about practical rationality. The structure of my argument is as follows. I begin by rejecting the view that rationalism is stronger than instrumentalism in being the logically stronger theory of the two (Section 2). I then focus on the possibility that rationalism involves logically stronger requirements than instrumentalism and show that instrumentalists and rationalists have reason to (and usually do) state their respective views such that such a logical asymmetry between the ought claims they each support does not obtain (Sections 3 and 4). Even if this were different, however, such an asymmetry would not entail an asymmetry in prima facie Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 justiﬁability (Section 5).

2. Rationalism is not logically stronger than instrumentalism

In the most straightforward sense, a theory T1 is stronger than a theory T2 if T1 entails T2, but not vice versa, that is, if T1 is logically stronger than T2 (consider, e.g. “the planets revolve around the sun on ellipses” vs. “the planets revolve around the sun”). Clearly, there is an interesting difference in strength between two theories logically related in this way. One reason is that the difference in logical strength might lead to a difference in burdens of justification. It is plausible to assume that if one is justified in believing T1 and if one accepts T2 on the basis of knowing that T1 entails T2, then one is also justified in believing T2, because of the facts that one is justified in believing T1 and that T1 7 entails, as one knows, T2. In this sense, one’s justificatory grounds for T1 ipso facto also justify T2. The converse is not true, if, as one knows, T2 does not entail T1.IfT1 and T2 are related in this way, then there is an asymmetry in burdens of justification since, in 20 Thomas Schmidt

order to justify T1, one needs to go beyond the justificatory grounds of T2, but not vice versa.8 One might think that rationalism is logically stronger than instrumentalism in just this sense. This is so since rationalists who take up the instrumental principle and add something to it in formulating their position, at least at first sight can be described as continuing past the point at which instrumentalists stop and, thereby, as going beyond instrumentalism, logically speaking. This sort of stopping, however, is not logically innocent. Crucially, instrumentalism involves a commitment not only to (IP) as a principle saying what one ought to do, but also to the view that all requirements of practical reason can be derived from (IP). The instrumentalist’s stopping at the point at which the rationalist continues, therefore, involves adding, over and above (IP), a “that’s it”-clause to her view. And it is the instrumentalist’s “that’s it”-clause and not, or at least not necessarily (more on this qualification in Sections 3 and 4), the instrumental principle, which rationalists deny. This is why rationalism does not go beyond instrumentalism but rather contradicts it. Therefore, rationalism is not logically stronger than instrumentalism. It thus turns out that even though it might seem that certain forms of rationalism say more than instrumentalism, this is not so if “saying more” is understood as “logically going beyond”. Rationalism, however, does seem to say more than instrumentalism in another sense of “saying more”: if rationalists endorse the instrumental principle in the same form as instrumentalists, and if rationalists add to this instrumental principle additional putative requirements about the normative acceptability of ends, then the ought claims stated by rationalism logically entail those involved in instrumentalism (in the sense of “entail” specified by standard propositional logic). Even though rationalism, then, is not the logically stronger theory, it might, therefore, be the theory involving the logically stronger requirements.9 Prima facie, it is not implausible that this might lead to an asymmetry between the two types of views that has implications for matters of justification. In Sections 3 and 4, I discuss whether rationalism really involves logically stronger requirements than instrumentalism. As it turns out, this would only be so if rationalists and instrumentalists stated their respective views in ways that they have independent reasons to avoid. In Section 5, I complete my defence of rationalism against the charge of bearing the burden of proof by showing that even under the assumption that rationalism states, or entails, the existence of logically stronger requirements than instrumentalism, this Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 would not entail that rationalism is more ambitious than instrumentalism in a way that makes instrumentalism the default view.

3. Wide scope vs. narrow scope In order for rationalism to involve logically stronger requirements than instrumentalism, it needs to be the case that rationalists accept the instrumental principle and that both rationalists and instrumentalists interpret it in the same way. Only then can rationalism be seen as taking up the instrumental principle in the form in which it is also accepted by instrumentalists and then conjoining further requirements to it. In this and in the next section, I show why rationalists and instrumentalists have reason to understand the instrumental principle in different ways, with the consequence that the requirements stated, or entailed, by rationalism are not logically stronger than those stated, or entailed, by instrumentalism. As is well known, there are two fundamentally different interpretations of the instrumental principle. Under the narrow-scope reading, the content of (IP) is given by Instrumentalism About Practical Reason: Not by Default 21

(IPN) if one has an end, then one ought to take the suitable means.

Under the wide-scope interpretation, (IP) is to be understood as saying that

(IPW) one ought to see to it that, if one has an end, then one takes the suitable means.

When it is interpreted in the narrow-scope sense, the instrumental principle speciﬁes one, and only one, option that one ought to take in order to satisfy it. When understood in the wide-scope sense, however, there are, as far as the instrumental principle is concerned, two ways of going about when one has a certain end: one can either take the suitable means, or else abandon the end.10 Instrumentalists typically adopt a narrow-scope interpretation of the instrumental principle, since (IPN), unlike (IPW), is consistent with, and, indeed, a way of expressing, the standard instrumentalist idea that all requirements of practical reason are grounded in individual ends or, given the way in which instrumentalists often ﬂesh out the notion of an end, in individual desires.11 Rationalists who want to stick to the instrumental principle, in contrast, have reason to adopt a wide-scope reading of it. For one thing, they are often concerned about what Bratman (1987, 24–27) has called the “bootstrapping problem”: given a narrow-scope interpretation of the instrumental principle, you can “bootstrap” the fact that you ought, or have a reason, to do something into existence just by adopting some end. For another thing, rationalists typically regard the instrumental principle as being in danger of entailing normatively implausible conclusions, since instrumentalists operate with a normatively unconstrained notion of those ends that give rise to oughts, or reasons.12 This is precisely the reason why rationalists bring into play criteria for the normative acceptability of ends. Doing so, however, is only consistent with the instrumental principle if it is understood in its wide-scope version. This is so since (IPW) only requires means-end-coherence. (IPN), by contrast, commits one to regard any ends which people actually have as giving rise to oughts, or reasons. This is why, when faced with the choice between the narrow-scope and the wide-scope versions of the instrumental principle, rationalists should reject (IPN) and adopt (IPW). Since (IPW) does not entail (IPN) and since there are ought claims entailed by (IPN) that are not entailed by (IPW), it follows that versions of rationalism subscribing to (IPW) and rejecting (IP ) do not involve logically stronger requirements than the standard version of instru-

Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 N mentalism, that is, instrumentalism subscribing to (IPN). And if this is so, then neither is rationalism logically stronger than instrumentalism, nor does rationalism entail ought claims that are logically stronger than those proclaimed by instrumentalism.

4. Rationalism without the instrumental principle Rationalism’s requirements are not logically stronger than instrumentalism’s requirements if rationalism involves a wide-scope instrumental principle (IPW) and instrumentalism adopts the instrumental principle in its narrow-scope version (IPN). However, even if instrumentalists adopted (IPW), the requirements entailed by such a form of instrumentalism would not be necessarily entailed by rationalism’s requirements, since it is an open question whether rationalists should decide to incorporate the instrumental principle into their theory at all. There are two main considerations that rationalists might see as speaking against accept- ing the instrumental principle. For one thing, it has been doubted that a wide-scope instrumental principle really avoids the problem of licensing inferences to normatively 22 Thomas Schmidt

implausible ought claims.13 For another thing, taking up a line of thought from Kolodny (2005, 542–547), one can doubt that the instrumental principle can be given a compelling normative rationale anyway. Kolodny’s challenge focuses on requirements of rationality that are formulated in a way which makes it an open question whether rationality is normative, that is, whether there are reasons to satisfy the requirements of rationality. Even so, the core of Kolodny’s challenge can be rephrased so that it also applies to normative principles like the ones discussed in this paper: how can the instrumental principle, in its wide-scope or in its narrow-scope interpretation, be justified? What is the normative rationale of (different forms of) the instrumental principle? If all requirements of practical reason can be seen as grounded in the instrumental principle, then there is – at least at first sight – an attractive option for answering these ques- tions. The point of acting rationally, one is then able to say, is to pursue what one cares about as well as possible.14 This idea, however, cannot be used to back up the wide-scope version within rationalism. Under the wide-scope reading, the instrumental principle does not call for optimally pursuing what one cares about, but for seeing to it that what one cares about and what one pursues fit each other. It is not obvious why this idea should be part of the rationalist’s normative package. This is why there are reasons why rationalists might want to abandon the instrumental principle and follow Raz in believing that “there is no distinctive form of rationality or of normativity that merits the name instrumental rationality or normativity” (Raz 2005a, 24). Given that rationalists, as their position has been introduced above, have substantial criteria normatively constraining ends, they might, instead of the instrumental principle, want to take on board a different normative principle, such as, for example,

(P) if one has a rationally permissible end, then one ought to take the suitable means,

or a transmission principle such as:

(T) if one ought to f, and c-ing is a necessary means for one to f, then one ought to c.

Arguably, these principles do not license bootstrapping of normatively unacceptable ought claims; there is reason to be conﬁdent that they can be given a normative rationale within a rationalist framework; and they can be seen as accounting for at least some of the intuitions Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 underlying the instrumental principle, even though they differ from it substantially.15 Despite this, there are also reasons against abandoning the instrumental principle and replacing it, for example, by (P) or (T) within a rationalist framework. Having the instrumental principle (under whatever interpretation) on board allows one, as Wallace (2001) has pointed out, to account for the at least prima facie plausible view that there is a normatively relevant distinction between agents who are strategically clever in the pursuit of their ends and agents who are not – no matter whether their ends are rationally permissible. With a version of the instrumental principle at hand, the rationalist can say that the clever person satisﬁes the requirements of instrumental rationality, while the stupid person does not. When one does away with the instrumental principle, one runs the risk of not being able to account for the phenomenon of cleverness within one’s theory about what one ought to do. Principles such as (P) and (T) are, with regard to accounting for cleverness, not suitable substitutes for (IP). This point can be illustrated by considering how four different characters relate to these principles. Stupid Killer (SK) is an individual intending to kill a victim but not undertaking Instrumentalism About Practical Reason: Not by Default 23

any steps towards this end. Rather, everything SK does enhances other people’s well-being – as opposed to Clever Killer (CK), who has the same end as SK, but does take the means suitable for achieving this end. Suppose that SK’s and CK’s ends are ones which, according to rationalism, one ought not to have.16 Stupid Philanthropist (SP), on the other hand, has an end that he ought to have, but fails to take the suitable means. In this, he differs from Clever Philanthropist (CP), who has the same ends as SP. Incidentally, the means intended by CP are exactly the ones which SK adopts. Many will have the intuition that there is, in terms of how one rationally ought to be, a difference between SK and CK. SK seems to fail to be as he ought to be in a way that CK does not. Subscribing to the instrumental principle (no matter whether in its narrow-scope or in its wide-scope version) enables one to account for this intuition: SK violates (IP), and CK satisﬁes it. If rationalists give up (IP), then they seem to deprive themselves of the resources needed for explaining the intuition. Adopting (P) or (T) instead of the instrumental principle does not help: since SK intends the means which he ought to take (given the ends which he ought to, but in fact does not, pursue), he violates neither (P) nor (T), while CK does. Analogous remarks apply to the intuition that at least one of the respects in which SK is not as he ought to be is also a respect in which SP is not as he ought to be. With the instrumental principle at hand, this idea can be readily explained, since both violate (IP). Adher- ents of (P) or (T), however, cannot account for it on the basis of these principles alone, since SK violates neither (P) nor (T), whereas SP violates both principles.17 These considerations conﬁrm the suspicion formulated above. If a theory of practical reason involves some version of the instrumental principle, then it is able to account for a number of intuitions about normative failure that seem to be straightforward instances of instrumental irrationality. Principles such as (P) or (T), which can be seen as distant rela- tives of the instrumental principle, do not in themselves provide an adequate substitute when it comes to accounting for the intuitions in question. Flatly denying that, intuitively, there is a respect in which SK, as opposed to CK, is not as he ought to be and that, again: intuitively, there is a respect in which both SK and SP are not as they ought to be, does not seem to be an option. Therefore, and since (P) and (T) are, with regard to accounting for cleverness, not suitable substitutes for the instrumental principle, rationalists rejecting the instrumental principle in any of its forms should think about alternative ways of accounting for the intuitions about normative failure seemingly support- ing it. Even though there are theoretical options worth considering,18 and despite the Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 reasons rationalists might see as speaking against the instrumental principle, there is something to be said for incorporating it into a rationalist framework. This is why even though the requirements stated, or entailed, by rationalism do not entail the requirements stated, or entailed, by forms of instrumentalism involving a narrow-scope instrumental principle, there might be plausible forms of rationalism (involving the instrumental principle) whose requirements do entail the requirements of certain forms of instrumentalism (namely versions of instrumentalism that accept a wide-scope interpretation of the instrumental principle).

5. Requirements and permissions Suppose, then, that rationalism’s requirements did entail instrumentalism’s requirements. As I argue in this section, this would not entail an asymmetry in burdens of justiﬁcation. Recall that rationalism and instrumentalism should be understood as involving the claim that actions that are neither required nor prohibited by reason are rationally permitted. 24 Thomas Schmidt

It then follows that if rationalism were to involve logically stronger requirements than instrumentalism, instrumentalism would involve logically stronger permissions. Therefore, in order for there to be an asymmetry with regard to burdens of justification between rationalism and instrumentalism that is grounded in the fact that rationalism involves logically stronger requirements, it would have to be the case that, ceteris paribus, requirements are in greater need of justification than permissions. But why should this be so? I discuss, and reject, three different ways one might try to ground a difference in justifiability in the logical asymmetry just described. First, one might point to the fact that, if rationalism and instrumentalism are logically related in the way just assumed, then there is a sense in which rationalism is psychologically more demanding than instrumentalism. Satisfying rationalism’s requirements would typically require greater cognitive awareness, or motivational efforts, or both, than satisfying instrumentalism’s requirements. Given that instrumentalism is, by contrast, more permiss- ive, it is generally easier to live up instrumentalism’s requirements. Such a difference in demandingness, however, has no implications for burdens of justification. Imagine a situation in which somebody asks for advice about what he ought to do, and gets different answers from a rationalist and an instrumentalist. Suppose further that following the rationalist’s advice would be motivationally more demanding than following the instrumentalist’s advice. Would this mere fact be a reason speaking in favour of the greater adequacy, or cor- rectness, of the instrumentalist’s advice? Of course not. Second, one might believe that normative theories involving ought claims demand things from individuals in another sense, roughly in the same sense as, for instance, political authorities demand things. Then, rationalism would interfere with individuals and, in particular, with individual freedom to a greater extent than instrumentalism, which would seem to entail a difference in burdens of justification. The premise of this argument, however, is false. Normative theories and, in particular, rationalism and instrumentalism do not in this sense demand, and they do not in this sense interfere with individuals and their freedom. Rationalism and instrumentalism are theories whose content consists in statements about what agents ought, or have reason to do. That A ought to f does not entail that anybody has the right to demand from A that he f, much less the permissibility of coercive measures with the aim of getting A to f. Letting A know that he ought to f need not involve a demand, but is normally just an assertion. Third, it is plausible to assume that if one has no indication at all as to how one ought to act in a particular situation, then it makes sense to act as if everything is permitted. One Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 might take this, in turn, to imply that, absent evidence as to how one ought to act, there is reason to believe that everything is permitted. If this were so, then there would be a sense in which permissions are in lesser need of justification than requirements. However, even if the normative theory stating that everything is permitted is practically equivalent with the stance that we might want to adopt absent evidences about requirements, the absence of such evidences does not provide any justification for the theory stating that everything is permitted. As far as this goes, therefore, there is no reason to believe that rationalists are in greater need of justifying their theory than instrumentalists if rationalism’s requirements are logically stronger than instrumentalism’s requirements.

6. Conclusion In this paper, I have argued that there are no structural differences between rationalism and instrumentalism in virtue of which rationalism is prima facie less defensible than Instrumentalism About Practical Reason: Not by Default 25

instrumentalism. Rationalism is not logically stronger than instrumentalism, and the requirements supported by rationalism do not entail the requirements supported by at least standard forms of instrumentalism. And even if this were not so, the resulting asymmetry would not have a philosophical upshot with regard to matters of justiﬁcation. As far as structural considerations go, instrumentalism, therefore, is by no means the default view in the theory of practical reason.

Acknowledgements In writing this paper, I have used material presented in talks at universities in Bayreuth, Berlin (HU), Konstanz, Leipzig, Mainz, St Andrews, Uppsala, and Vienna. I have greatly beneﬁted from the dis- cussions on these occasions. Also, I have proﬁted from stimulating exchanges about or comments on earlier sketches of material which ended up in this paper with Vuko Andric, Norbert Anwander, Krister Bykvist, Roger Crisp, Christoph Fehige, Tim Henning, Kent Hurtig, Geert Keil, Felix Koch, Kirsten Meyer, Wlodek Rabinowicz, Michael Smith, Folke Tersman, Holm Tetens, and Jay Wallace. I am particularly indebted to Simon Gaus, Jan Gertken, Benjamin Kiesewetter, and two anonymous referees for very helpful comments on earlier drafts.

Disclosure statement No potential conﬂict of interest was reported by the author.

Notes 1. Throughout this paper, the ought referred to is the ought of practical reason. I do not take a stance on whether, besides the ought of practical reason, there are other oughts governing actions and intentions. With regard to the argumentative purposes of this paper, it can be left open whether requirements of practical reason such as the instrumental principle need to be sup- plemented by conditions relativising the ought to the epistemic situation of the agent. 2. I have introduced the instrumental principle in a way that is open to a narrow-scope and a wide- scope interpretation. I take up the question which one of these is to be preferred in Section 3. 3. One should understand this form of rationalism as also involving the claim that all requirements of practical reason can be derived from the conjunction of (IP) and the additional criteria normatively constraining ends. These criteria are to be understood as “substantial” in the sense employed here if they do not relativise the acceptability of ends to other ends. – Introducing requirements that normatively constrain the choice of ends is sometimes motivated by the Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 desire to exclude immoral ends from the set of those the pursuit of which is permitted by reason and sometimes for including ends the pursuit of which reason requires. A famous example for the latter is Kant’s view that one’s own perfection and the happiness of others are, as he puts it, ends that are also duties, that is, ends which are prescribed by reason (Kant 1790, 386). A morally neutral example for a case in which somebody does not have an end which he rationally ought to have is Parfit’s (1984, 124) “Future-Tuesday-Indifference” (involving somebody who does not care about possible pains or pleasures that happen to him on future Tuesdays). For more examples, see Hooker and Streumer (2004, 67–69). 4. One might question whether Nozick really intends his term “instrumental rationality” to refer to instrumentalism in the way I have introduced it above. Alternatively, he might just have in mind the instrumental principle, and not the instrumentalist view that there is nothing more to practical reason than this principle. That he does indeed mean to refer to instrumentalism can be con- firmed with reference to other passages of his book. He writes, for example, that according to the “instrumental conception, rationality consists in the effective and efficient achievement of goals, ends, and desires” (Nozick 1993, 64). And in a footnote on the same page, he quotes, again with reference to the instrumental theory of rationality, a passage from H. Simon who maintains that “[r]eason is wholly instrumental. It cannot tell us where to go; at best it can tell us how to get there” (Nozick 1993, 64). 26 Thomas Schmidt

5. The view that to ask what is rational is to ask what one ought to do is explicitly stated, for instance, in Gibbard (1990, 6–7). 6. See, in particular, the discussion following Kolodny (2005). 7. The principle that justified belief is closed under entailment in all cases is not uncontroversial (for some points of discussion and further references, see Luper [2012, § 6.2]), but there is reason to believe that it is possible to weaken it so that it covers the case under discussion here without being vulnerable to counterexamples. 8. Or, to put the point more cautiously, this will be so at least in the general case. In specific cases of theories that are logically related in the described way the available justification for the logically weaker theory might, as a matter of fact, also justify the logically stronger theory. 9. This would be the case if the form of the ought claims supported by rationalism were O(A)& O(B) and if instrumentalism only involved the ought claim O(A) (or if in some other way the set of ought claims supported by instrumentalism turns out to be a proper subset of the set of ought claims supported by rationalism). 10. In the recent literature, the importance of the distinction between wide-scope and narrow-scope interpretations of requirements of rationality has been stressed, among many others, by Broome (1999, 2007) and Wallace (2001). 11. Important examples for such instrumentalist views are the ones defended by Fehige (2001) and Schroeder (2007). 12. This problem seems to be particularly pressing for a version of instrumentalism such as Schroe- der’s, according to which every desire gives rise to a corresponding reason. Schroeder attempts to solve this, as he calls it, “too many reasons problem” by rejecting what he calls proportion- alism, that is, the view that “when a reason is explained by a desire [ ...], its weight varies in proportion to the strength of that desire” (Schroeder 2007, 98). 13. See, in particular, Raz (2005a, 12) and Setiya (2007, 656). 14. This ties in nicely with Fehige’s description of instrumentalism as being the attempt to spell out what he calls the “Hearty View”, the core of which is formed by the following two claims: “Some things are dear to our hearts. To act rationally [ ...] means in essence: to look after these things, as best as we can” (Fehige 2001, 49). 15. Principles such as these are, at least on the face of it, not without problems that would need to be discussed in greater detail than it is possible in this paper. For instance, it seems that you violate a requirement implied by (P) if you abandon a permissible end and not take the necessary means. (T), on the other hand, is applicable only in cases in which one ought to adopt certain ends and does not imply anything for cases of ends that are merely permissible. For a defence of (T) against objections that have been raised in the recent literature, see Kiesewetter (2015). 16. Nothing depends on the ends of the killers being morally dubious. Other, nonmoral examples would do the same job (see note 3 above). 17. The following table illustrates how the four individuals introduced above violate (“2”) or satisfy (“+”) these principles (I understand the notion of satisfying a principle to be applicable

Downloaded by [Thomas Schmidt] at 00:03 06 April 2016 also in cases where the antecedent of the principle’s conditional is false):

SK CK SP CP Instrumentalism (IP) – + – + Rationalism Has permissible ends – – ++ (P), (T) + –– +

18. Raz, for example, has suggested that if somebody fails to take a means to an end, this does not ipso facto mean that he is in a state which he has a reason to avoid, but rather “that he is not functioning properly” (Raz 2005b, 10), that is, that his rational capacity (the point of which according to Raz is to enable him to respond to reasons) is impaired in some way or another. Raz’s “myth view” is helpfully characterised, taken up and extended in Kolodny (2008).

Notes on contributor Thomas Schmidt is Professor of Philosophy at Humboldt-Universita¨t zu Berlin. His current research focuses on issues in normative ethics, moral epistemology, and the theory of normativity. Instrumentalism About Practical Reason: Not by Default 27

References Bratman, M. E. 1987. Intention, Plans, and Practical Reason. Cambridge, MA: Harvard University Press. Broome, J. 1999. “Normative Requirements.” Ratio 12: 398–419. Broome, J. 2007. “Wide or Narrow Scope?” Mind 116: 359–370. Dreier, J. 2001. “Humean Doubts about Categorical Imperatives.” In Varieties of Practical Reasoning, edited by E. Millgram, 27–47. Cambridge, MA: MIT Press. Fehige, Ch. 2001. “Instrumentalism.” In Varieties of Practical Reasoning, edited by E. Millgram, 49– 76. Cambridge, MA: MIT Press. Gibbard, A. 1990. Wise Choices, Apt Feelings: A Theory of Normative Judgment. Cambridge, MA: Harvard University Press. Hampton, J. E. 1998. The Authority of Reason. Cambridge: Cambridge University Press. Hooker, B., and B. Streumer. 2004. “Procedural and Substantive Practical Rationality.” In The Oxford Handbook of Rationality, edited by A. R. Mele, and P. Rawling, 57–74. New York: Oxford University Press. Kant, I. 1790. The Metaphysics of Morals. Translated by M. J. Gregor. Cambridge: Cambridge University Press 1991 (Ak. Ed. Vol. IV). Kiesewetter, B. 2015. “Instrumental Normativity: In Defense of the Transmission Principle.” Ethics 125: 921–946. Kolodny, N. 2005. “Why Be Rational?” Mind 114: 509–563. Kolodny, N. 2008. “The Myth of Practical Consistency.” European Journal of Philosophy 16: 366– 402. Korsgaard, C. M. 1986. “Skepticism about Practical Reason.” The Journal of Philosophy 83: 5–25. Luper, S. 2012. “The Epistemic Closure Principle.” In The Stanford Encyclopedia of Philosophy (Fall 2012 Edition), edited by E. N. Zalta. ,http://plato.stanford.edu/archives/fall2012/entries/closure- epistemic/.. Mackie, J. L. 1977. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin. Nozick, R. 1993. The Nature of Rationality. Princeton, NJ: Princeton University Press. Parﬁt, D. 1984. Reasons and Persons. Oxford: Oxford University Press. Raz, J. 2005a. “The Myth of Instrumental Rationality.” Journal of Ethics and Social Philosophy 1. Raz, J. 2005b. “Instrumental Rationality: A Reprise.” Journal of Ethics and Social Philosophy 1, Symposium I. Schroeder, M. 2007. Slaves of the Passions. Oxford: Oxford University Press. Setiya, K. 2007. “Cognitivism About Instrumental Reason.” Ethics 117: 649–673. Wallace, R. J. 2001. “Normativity, Commitment, and Instrumental Reason.” Philosophers’ Imprint 1: 1–26. Williams, B. (1979) 1981. “Internal and External Reasons.” Chap. 8 in Moral Luck. Reprint, Cambridge: Cambridge University Press. Downloaded by [Thomas Schmidt] at 00:03 06 April 2016